On 18 June 2014 16:05, Dave <[email protected]> wrote: > On Tue, Jun 17, 2014 at 10:55 PM, Jake Farrell <[email protected]> > wrote: > >> Hey everyone >> Sorry I had to leave early from this meeting, wanted to follow up on the >> points that must take place >> >> * The ASF is the canonical location for source code >> > > We all agree that ASF Git is our canonical repo, and that is the source > from which will will create our releases. That is what our existing process > specifies. In the IRC meeting we agreed to also put a clear notification on > the GitHub repo that it is not the canonical repo. > > > * There must be a clear audit trail for the contribution being committed >> > > Our existing process covers this, but we need complete the work of echoing > all GitHub PR comments and actions to our dev or commits mailing list. > > > >> * The commit must occur to ASF hardware by a committer, not to Github >> or via a sync process >> > > Can you please provide a link to the ASF policy that specifies this rule > because the above sentence does not make sense to me. Commits from an > incoming GitHub PR *always* occur on GitHub. > > In our process the commit that merges a PR also happens on GitHub but the > commit that merges the PR into ASF Git happens on ASF Git, from a > committer. > > Where is the rule that says a sync process cannot be used? > if the committer is clearly identified and not hidden by some git common user, I cannot see sync as a problem.
BUT if sync happens with one user, and not the original committer we loose traceability and that would break our policy. rgds jan I. > > > >> * All releases must be cut from the source located at the ASF >> > > Definitely. That is what our policy says. > > Thanks for helping us work this out. > > - Dave > > > > > >> >> On Tue, Jun 17, 2014 at 5:31 PM, ASF IRC Bot <[email protected] >> > >> wrote: >> >> > Summary of IRC Meeting in #usergrid at Tue Jun 17 20:12:03 2014: >> > >> > Attendees: snoopdave, rockerst_, jfarrell, Humbedooh, toddnine, >> > rockerston, sfeldman, lmcgibbn >> > >> > - Preface >> > - usergrid git workflow >> > >> > >> > IRC log follows: >> > >> > ## Preface ## >> > ## usergrid git workflow ## >> > [Tue Jun 17 20:12:48 2014] <jfarrell>: everyone involved in todays >> meeting >> > can we please do a quick roll call >> > [Tue Jun 17 20:12:58 2014] <jfarrell>: here >> > [Tue Jun 17 20:12:59 2014] <snoopdave>: snoopdave - Dave Johnson >> > [Tue Jun 17 20:13:26 2014] <rockerston>: rockerston - Rod Simpson >> > [Tue Jun 17 20:13:36 2014] <toddnine>: toddnine - Todd Nine (cause I’m >> > creative like that) >> > [Tue Jun 17 20:13:38 2014] <sfeldman>: sfeldman - Shawn Feldman >> > [Tue Jun 17 20:14:13 2014] <lmcgibbn>: lmcgibbn - Lewis John McGibbney >> > [Tue Jun 17 20:15:17 2014] <lmcgibbn>: afternoon team >> > [Tue Jun 17 20:15:17 2014] <snoopdave>: ok, so we came up with a >> > contribution work flow and voted it in >> > [Tue Jun 17 20:15:36 2014] <snoopdave>: then we got word that it was not >> > appropriate, but we don’t know exactly why >> > [Tue Jun 17 20:15:43 2014] <rockerston>: hey, hey, lewis! >> > [Tue Jun 17 20:15:49 2014] <jfarrell>: currently the workflow that is >> > being used was brought up on board@ and infra@ and my initial email was >> > to start the conversation around how we can become in compliance with >> > existing policies >> > [Tue Jun 17 20:16:09 2014] <snoopdave>: cool. So what specific polcies >> are >> > we violating? >> > [Tue Jun 17 20:16:21 2014] <jfarrell>: this was triggered by the sync >> you >> > had running on people.a.o >> > [Tue Jun 17 20:16:36 2014] <jfarrell>: the commits can not occur at >> > github, they much occur against git-wip >> > [Tue Jun 17 20:17:02 2014] <snoopdave>: right. My sync process was >> judged >> > to be insecure becuase I was storing my ASF creds in a .netrc file (only >> > readable by me) >> > [Tue Jun 17 20:17:04 2014] <lmcgibbn>: For those who need to look >> through >> > current documentation >> > [Tue Jun 17 20:17:04 2014] <lmcgibbn>: >> > >> https://cwiki.apache.org/confluence/display/usergrid/Contributor+Workflow+Policy >> > [Tue Jun 17 20:17:19 2014] <snoopdave>: that is a fair criticism >> > [Tue Jun 17 20:18:11 2014] <snoopdave>: but is there a policy that >> > prohibits that type of sync from GitHub to ASF Git? >> > [Tue Jun 17 20:18:34 2014] <rockerston>: where does it say that commits >> > can't be done against github? >> > [Tue Jun 17 20:18:38 2014] <rockerston>: and why? >> > [Tue Jun 17 20:18:39 2014] <jfarrell>: yes, as the commit is not >> occurring >> > to the asf, it is occuring to github >> > [Tue Jun 17 20:18:46 2014] <rockerston>: why is that bad? >> > [Tue Jun 17 20:19:03 2014] <rockerston>: git uses a peer to peer >> structure >> > [Tue Jun 17 20:19:15 2014] <rockerston>: there is no canonical source >> > [Tue Jun 17 20:19:27 2014] <rockerston>: they are clones >> > [Tue Jun 17 20:21:12 2014] <toddnine>: Ultimately git is a P2P source >> > control system. There is no canonical source for a SHA, it can >> originate >> > from any system, even a local repo. As long as there is a log of all >> sha’s >> > commited in the Apache repository, why does it matter where it >> originates >> > from? >> > [Tue Jun 17 20:21:20 2014] <jfarrell>: yes, and that gives it even more >> > reason to have the commit occur at the asf and not have it synced >> > [Tue Jun 17 20:21:50 2014] <jfarrell>: becuase you can rewite the >> > author/committer and then sync it over, or in the case of how it was >> > occurring all commits@ had snoopdave as the origonator >> > [Tue Jun 17 20:21:57 2014] <jfarrell>: as committers we need to ensure >> > that the asf policies around controbutions are being met and when we >> make a >> > commit to the repo we are verifying that all these policies have been >> met >> > and ip clearance is met >> > [Tue Jun 17 20:22:07 2014] <lmcgibbn>: My understanding is that the code >> > contained at git-wip repos is the official repos >> > [Tue Jun 17 20:22:09 2014] <snoopdave>: also, in the JClouds process >> then >> > manually bring in commits that occured at GitHub and cause them to be >> > pushed to ASF Git — I don’t see the difference — the contributor’s >> commits >> > originate on GitHub them get pushed to ASF Git >> > [Tue Jun 17 20:22:18 2014] <jfarrell>: if that is not the case for a >> given >> > commit then we need to remove it and act accordingly, this is the >> primary >> > function of a PMC >> > [Tue Jun 17 20:22:48 2014] <jfarrell>: jclouds repo does not sync >> directly >> > to the ASF >> > [Tue Jun 17 20:23:04 2014] <jfarrell>: the committer is responsible for >> > making the commit and pushing it over to git-wip >> > [Tue Jun 17 20:23:30 2014] <snoopdave>: there are two types of commits >> in >> > this discussion, the contributor’s commits that happen at GitHub — and >> the >> > accepting committer’s commit that merges the change >> > [Tue Jun 17 20:23:39 2014] <toddnine>: I’m not convinced our history is >> > overwritten with this process. Observe this history. >> > https://github.com/apache/incubator-usergrid/commits/two-dot-o?page=2 >> > [Tue Jun 17 20:23:46 2014] <jfarrell>: jclouds was also setup before all >> > the github integration work we did, they are looking at whats needed to >> > switch over and use it over their current workflow >> > [Tue Jun 17 20:24:25 2014] <toddnine>: Also, isn’t it the >> responsibility >> > of the committer to ensure the contributer has an ICLA on file before >> > merging, regardless of the medium? >> > [Tue Jun 17 20:24:44 2014] <lmcgibbn>: Although the usergrid workflow >> was >> > modeled largely on the jClouds one... I am not sure if us quoting them >> > drives on the Usergrid agenda. >> > [Tue Jun 17 20:33:00 2014] <lmcgibbn>: toddnine: I am not sure if it is >> > the committer's responsibility >> > [Tue Jun 17 20:33:00 2014] <lmcgibbn>: Unless stated in the contribution >> > [Tue Jun 17 20:33:00 2014] <lmcgibbn>: then >> > [Tue Jun 17 20:33:00 2014] <lmcgibbn>: *everything* >> > [Tue Jun 17 20:33:03 2014] <snoopdave>: my main problem with the >> workflow >> > you’ve proposed Jake is additonal manual work to accept a commit from a >> > contributor, and the additonal work a committer must do do get his work >> > back to ASF Git (since we use GitHub as our code review system for all >> > commits) >> > [Tue Jun 17 20:36:31 2014] <rockerst_>: right now we don't have the job >> > turned on >> > [Tue Jun 17 20:36:36 2014] <lmcgibbn>: ... that's if there is much to >> > actually fix >> > [Tue Jun 17 20:37:11 2014] <rockerst_>: we just need to set up the >> > webhooks in github >> > [Tue Jun 17 20:37:22 2014] <rockerst_>: so the notifications go out to >> the >> > appropriate ML >> > [Tue Jun 17 20:37:38 2014] <snoopdave>: seems like we don’t have and >> > answer to what is allowed and disallowed by ASF policy. Is there some >> > specific person who raised issues with our process? somebody who we >> need to >> > convince? somebody who is “the decider” in our case? >> > [Tue Jun 17 20:38:25 2014] <lmcgibbn>: snoopdave: who made the >> > determination to revoke your account? >> > [Tue Jun 17 20:38:31 2014] <lmcgibbn>: Maybe we can start there. >> > [Tue Jun 17 20:38:46 2014] <jfarrell>: the other concern was that >> > github.com/usergrid is not asf owned and if commits are occuring there >> > then there is no security prevention in place (like ldap with git-wip) >> > [Tue Jun 17 20:39:47 2014] <rockerst_>: you have to have a GH account, >> > which is secure. >> > [Tue Jun 17 20:39:55 2014] <rockerst_>: what is not secure? >> > [Tue Jun 17 20:40:19 2014] <jfarrell>: gh account != apache account, can >> > add non committer to the usergrid org and they can make commits >> > [Tue Jun 17 20:40:45 2014] <rockerst_>: but isn't that our >> responsibility? >> > [Tue Jun 17 20:41:00 2014] <rockerst_>: to make sure no unauthorized >> > persons are added? >> > [Tue Jun 17 20:41:23 2014] <lmcgibbn>: Yes I suppose it is... >> > [Tue Jun 17 20:41:45 2014] <lmcgibbn>: and it is also our job as >> > incubating mentors to make sure that this is the case as well. >> > [Tue Jun 17 20:41:59 2014] <snoopdave>: just like it is our >> responsbility >> > to ensure people submit contributions under ASF and submit ICLAs when >> > necessary >> > [Tue Jun 17 20:42:13 2014] <lmcgibbn>: I think everyone is good on this >> > front... there are no problems here. >> > [Tue Jun 17 20:42:58 2014] <jfarrell>: yes, and using the ASF resources >> we >> > have done so thus far, as far as github.com/usergrid as a mentor I have >> > no access to it >> > [Tue Jun 17 20:43:17 2014] <lmcgibbn>: Neither do I >> > [Tue Jun 17 20:43:30 2014] <toddnine>: Well that’s easily recified :) >> > [Tue Jun 17 20:43:31 2014] <jfarrell>: i dont know what other mentors >> > do/dont have access, but know that in the ASF everyone has the correct >> > permissions >> > [Tue Jun 17 20:43:37 2014] <toddnine>: Usernames? >> > [Tue Jun 17 20:44:12 2014] <lmcgibbn>: toddnine: usernames for Github? >> > [Tue Jun 17 20:44:22 2014] <toddnine>: lmcgibbn: Correct >> > [Tue Jun 17 20:44:33 2014] <lmcgibbn>: My name is lewismc >> > [Tue Jun 17 20:44:34 2014] <jfarrell>: if there is something preventing >> us >> > from using the asf resources then I would love to work to add them >> > [Tue Jun 17 20:45:09 2014] <jfarrell>: i am only aware of jclouds as the >> > only other external org within github >> > [Tue Jun 17 20:45:50 2014] <toddnine>: Nothing is *wrong* with ASF per >> se >> > right >> > [Tue Jun 17 20:46:03 2014] <snoopdave>: the issue is that the project >> > wants to use GH for all code reviews and wants to hook into the huge >> number >> > of potential contributors using GitHub >> > [Tue Jun 17 20:46:09 2014] <rockerst_>: the reason that we want to use >> > github is because it is very user friendly, allows us do do awesome code >> > reviews, gives us access to a massive community of developers... >> > [Tue Jun 17 20:46:23 2014] <snoopdave>: and I don’t think those desires >> > are incompatible with ASF >> > [Tue Jun 17 20:46:28 2014] <toddnine>: It’s simply that GH is far more >> > popular. Our goal is to attract quality contributors. Creating >> > artificially imposed barriers to this seems to counter the Apache >> spirit. >> > [Tue Jun 17 20:46:47 2014] <jfarrell>: which can be done using >> > github/apache/incubator-usergrid, no? >> > [Tue Jun 17 20:47:10 2014] <rockerst_>: Jake, we can't do commits and >> > accept PRs there >> > [Tue Jun 17 20:47:17 2014] <rockerst_>: it is a read-only mirror >> > [Tue Jun 17 20:47:33 2014] <jfarrell>: >> > https://help.github.com/articles/closing-issues-via-commit-messages >> > [Tue Jun 17 20:47:34 2014] <snoopdave>: like I said before “my main >> > problem with the workflow you’ve proposed Jake is additonal manual work >> to >> > accept a commit from a contributor, and the additonal work a committer >> must >> > do do get his work back to ASF Git (since we use GitHub as our code >> review >> > system for all commits)” >> > [Tue Jun 17 20:47:48 2014] <rockerst_>: also you can't restrict access >> to >> > only /apache/incubator-usergrid >> > [Tue Jun 17 20:47:59 2014] <toddnine>: lmcgibbn: You’re gtg on the >> > usergrid/usergrid >> > [Tue Jun 17 20:48:35 2014] <jfarrell>: git-wip takes the initial commit >> -> >> > syncs to the official mirror on git.a.o and then github picks up this >> > mirror and closed any pr's based on the commit hash or commit message >> > [Tue Jun 17 20:48:37 2014] <rockerst_>: also, what is the difference >> > between using github/apache and usgin github/usergrid >> > [Tue Jun 17 20:48:38 2014] <lmcgibbn>: thank you toddnine >> > [Tue Jun 17 20:51:27 2014] <lmcgibbn>: folks >> > [Tue Jun 17 20:51:33 2014] <lmcgibbn>: AFAICT >> > [Tue Jun 17 20:52:03 2014] <lmcgibbn>: The document we maintain doe not >> > cover contributions and commits to the 2-dot-0 branch >> > [Tue Jun 17 20:52:06 2014] <lmcgibbn>: Is this correct? >> > [Tue Jun 17 20:53:07 2014] <snoopdave>: I thought we agreed that all >> code >> > that goes into master must be PR’d and reviewd, but code that goes into >> > branches does not (until it is merged with master) >> > [Tue Jun 17 20:53:41 2014] <lmcgibbn>: snoopdave: thank you for >> clarifying >> > [Tue Jun 17 20:53:47 2014] <lmcgibbn>: That sounds logical to me >> > [Tue Jun 17 20:54:12 2014] <lmcgibbn>: It also prevent barrier to entry >> > for other wanting to work on branches/issues. >> > [Tue Jun 17 20:54:56 2014] <lmcgibbn>: Can I suggest that in an attempt >> to >> > converge our thoughts on this topic we attempt to address the following >> 4 >> > points >> > [Tue Jun 17 20:55:13 2014] <lmcgibbn>: 1. Pull request created: includes >> > the date, time, username, description of the pull request and link. >> > [Tue Jun 17 20:55:22 2014] <lmcgibbn>: 2. Pull request commented: >> includes >> > the date, time, username, comment content and link >> > [Tue Jun 17 20:55:31 2014] <lmcgibbn>: 3. Pull request merge: includes >> the >> > date, time, username, comment content and link >> > [Tue Jun 17 20:55:39 2014] <lmcgibbn>: 4. Pull request closed: includes >> > the date, time, username, comment content and link >> > [Tue Jun 17 20:56:13 2014] <lmcgibbn>: For all of the above we need to >> set >> > up a mechanism that causes every pull request to be recorded on the >> project >> > commits list as emails: >> > [Tue Jun 17 20:56:27 2014] <lmcgibbn>: Does anyone have an issue with >> the >> > above? >> > [Tue Jun 17 20:56:41 2014] <Humbedooh>: ACTION peeks.. >> > [Tue Jun 17 20:56:58 2014] <jfarrell>: i've got to run, i pinged >> > Humbedooh, who is one of the other git admins here at the ASF in >> adition to >> > being on the IPMC and root@ >> > [Tue Jun 17 20:56:58 2014] <toddnine>: lmcgibbn: Not at all. We simply >> > need to echo that into the ML right? >> > [Tue Jun 17 20:57:17 2014] <snoopdave>: thanks Jake >> > [Tue Jun 17 20:57:21 2014] <lmcgibbn>: jfarrell: thanks for your time >> > [Tue Jun 17 20:57:25 2014] <Humbedooh>: so...what's your gripe with >> using >> > a mirror on github? >> > [Tue Jun 17 20:57:28 2014] <lmcgibbn>: Humbedooh: hi hi >> > [Tue Jun 17 20:57:30 2014] <jfarrell>: Humbedooh: can you close out the >> > meeting when done so i can catch up on whats missed >> > [Tue Jun 17 20:57:45 2014] <Humbedooh>: not to be grumpy cat, but it >> works >> > well for just about every other ASF project :) >> > [Tue Jun 17 20:58:02 2014] <lmcgibbn>: toddnine: AFACS yes... lets see >> how >> > things pan out though >> > [Tue Jun 17 20:58:09 2014] <Humbedooh>: we have github->asf and >> > asf->github replication of code, messages etc >> > [Tue Jun 17 20:58:14 2014] <snoopdave>: like I said before “my main >> > problem with the workflow you’ve proposed Jake is additonal manual work >> to >> > accept a commit from a contributor, and the additonal work a committer >> must >> > do do get his work back to ASF Git (since we use GitHub as our code >> review >> > system for all commits)” >> > [Tue Jun 17 20:58:36 2014] <snoopdave>: the project wants to use GH for >> > all code reviews and wants to hook into the huge number of potential >> > contributors using GitHub >> > [Tue Jun 17 20:58:57 2014] <Humbedooh>: 1) so you spend 1 more minute, >> > that doesn't strike me as a big deal >> > [Tue Jun 17 20:59:06 2014] <Humbedooh>: 2) nobody is preventing you from >> > using github >> > [Tue Jun 17 20:59:21 2014] <Humbedooh>: there is no "secret bar" for >> > github users from their perspective >> > [Tue Jun 17 20:59:33 2014] <Humbedooh>: they add a PR, some code, and >> > that's still all they need to do >> > [Tue Jun 17 20:59:47 2014] <Humbedooh>: if it's a larger issue, they >> sign >> > an ICLA, as with all other projects >> > [Tue Jun 17 21:00:26 2014] <Humbedooh>: but the only difference between >> a >> > mirror and r/w on github is that you have to pull their changes, check >> it, >> > and send it to git-wip >> > [Tue Jun 17 21:00:33 2014] <Humbedooh>: which I would expect you to do >> > anyway >> > [Tue Jun 17 21:01:06 2014] <Humbedooh>: do note I use 'you' in plural >> > sense :) >> > [Tue Jun 17 21:01:19 2014] <rockerst_>: in RE: 1), it isn't 1 more >> minute. >> > it is an unpleasant process that doesn't make sense >> > [Tue Jun 17 21:01:52 2014] <snoopdave>: we need to do that 1 minute >> thing >> > for every push we do, that is significant >> > [Tue Jun 17 21:01:54 2014] <rockerst_>: in RE 2), not sure what you mean >> > there. We are using Github and doing so very effectively >> > [Tue Jun 17 21:02:04 2014] <rockerst_>: not sure what you mean by >> "secret >> > bar" >> > [Tue Jun 17 21:02:37 2014] <rockerst_>: We have come up with a ver >> > streamlined process >> > [Tue Jun 17 21:02:48 2014] <lmcgibbn>: Humbedooh: >> > [Tue Jun 17 21:02:56 2014] <rockerst_>: and nobody has been able to >> > effectively explain what the problems are >> > [Tue Jun 17 21:03:05 2014] <lmcgibbn>: Can you please review the 4 >> bullet >> > points I put above? >> > [Tue Jun 17 21:03:12 2014] <rockerst_>: the argument has only been "do >> it >> > this other way" >> > [Tue Jun 17 21:03:25 2014] <lmcgibbn>: The Usergrid incubating community >> > [Tue Jun 17 21:03:28 2014] <lmcgibbn>: VOTE'd >> > [Tue Jun 17 21:03:33 2014] <lmcgibbn>: to put in place >> > [Tue Jun 17 21:03:36 2014] <Humbedooh>: lmcgibbn: that's all taken care >> > off by the github->asf integration >> > [Tue Jun 17 21:03:36 2014] <lmcgibbn>: and then stick to >> > [Tue Jun 17 21:03:50 2014] <lmcgibbn>: the workflow which we are >> discussing >> > [Tue Jun 17 21:03:54 2014] <rockerst_>: what lewis proposes makes >> perfect >> > sense >> > [Tue Jun 17 21:03:58 2014] <Humbedooh>: whenever something happens on a >> > github mirror, it gets replicated on the ML >> > [Tue Jun 17 21:04:25 2014] <rockerst_>: we don't want to use this other >> > process >> > [Tue Jun 17 21:04:36 2014] <Humbedooh>: rockerst_: it was implied that >> > there was some issue with using a github mirror as opposed to a r/w >> > repository >> > [Tue Jun 17 21:04:40 2014] <rockerst_>: because it is cumbersome and >> kind >> > of sucky >> > [Tue Jun 17 21:04:49 2014] <Humbedooh>: otherwise, I don't know why >> there >> > was a need to mention that you want to use github >> > [Tue Jun 17 21:04:55 2014] <rockerst_>: so why can't we use our >> workflow? >> > [Tue Jun 17 21:05:16 2014] <Humbedooh>: a r/w repository? >> > [Tue Jun 17 21:05:19 2014] <rockerst_>: the discussion isn't about >> using a >> > different workflow >> > [Tue Jun 17 21:05:32 2014] <rockerst_>: it is about using the workflow >> > that we have laid out >> > [Tue Jun 17 21:05:43 2014] <rockerst_>: we were told that we are >> violating >> > apache policy >> > [Tue Jun 17 21:05:50 2014] <rockerst_>: but i have not seen any evidence >> > of that >> > [Tue Jun 17 21:06:17 2014] <rockerst_>: and nobody has been able to >> point >> > to anything that we can't comply with >> > [Tue Jun 17 21:06:38 2014] <Humbedooh>: if the canonical source is not >> on >> > ASF hardware and if the commits are not done by an ASF committer, then >> it's >> > not within policy >> > [Tue Jun 17 21:07:05 2014] <lmcgibbn>: The Usergrid workflow is >> completely >> > new to me. I am always accustomed to using git-wip for pushing code >> > changes. However the (diverse) community of developers has VOTE'd to use >> > it. It we can use the workflow then I would like to make best efforts to >> > meet this. >> > [Tue Jun 17 21:07:09 2014] <lmcgibbn>: @Humbedooh >> > [Tue Jun 17 21:07:23 2014] <lmcgibbn>: I think that the canonical source >> > is at ASF >> > [Tue Jun 17 21:07:27 2014] <lmcgibbn>: it is on git-wip >> > [Tue Jun 17 21:07:34 2014] <snoopdave>: policy only says that releases >> > must be made from canonical repo at the ASF and that is what we do >> > [Tue Jun 17 21:07:36 2014] <lmcgibbn>: This is where releases will be >> done >> > from >> > [Tue Jun 17 21:07:45 2014] <lmcgibbn>: on the second point >> > [Tue Jun 17 21:07:46 2014] <Humbedooh>: I'm struggling to see what the >> big >> > issue is here - is it _just_ that you want to be able to merge directly? >> > [Tue Jun 17 21:08:00 2014] <lmcgibbn>: commits will always be done by a >> > VOTE'd upon committer >> > [Tue Jun 17 21:08:06 2014] <rockerst_>: all of our commits are tracked >> > [Tue Jun 17 21:08:07 2014] <rockerst_>: yes >> > [Tue Jun 17 21:08:30 2014] <Humbedooh>: so you never just hit the >> 'merge' >> > button or whatever it's called? >> > [Tue Jun 17 21:08:31 2014] <lmcgibbn>: That is the job of the Incubating >> > PPMC (including mentors) as well as the PMC once the community and >> project >> > graduates. >> > [Tue Jun 17 21:08:42 2014] <toddnine>: Humbedooh: Correct. The extra >> > download/upload step gets time consuming with a lot of requests, >> especially >> > trivial contributions >> > [Tue Jun 17 21:09:12 2014] <toddnine>: If we can just review these small >> > contributions, and press Accept, it’s significantly easier for us. >> > [Tue Jun 17 21:09:13 2014] <rockerst_>: all code that goes to Master are >> > done with PRs >> > [Tue Jun 17 21:09:32 2014] <Humbedooh>: toddnine: I'm not sure you are >> > covered by the ASF if you do that >> > [Tue Jun 17 21:09:47 2014] <Humbedooh>: and I don't want your house to >> be >> > on the line if someone sues usergrid :) >> > [Tue Jun 17 21:10:00 2014] <lmcgibbn>: Humbedooh: >> > [Tue Jun 17 21:10:02 2014] <lmcgibbn>: can you clarify >> > [Tue Jun 17 21:10:10 2014] <snoopdave>: humbedooh: this is the workflow >> we >> > are following: >> > >> https://cwiki.apache.org/confluence/display/usergrid/Contributor+Workflow+Policy >> > [Tue Jun 17 21:10:38 2014] <lmcgibbn>: It was my understanding that all >> > code submitted via patch and/or pull request is licensed as ASLv2.0 >> > [Tue Jun 17 21:10:47 2014] <lmcgibbn>: Unless the author says otherwise >> > [Tue Jun 17 21:11:06 2014] <lmcgibbn>: This was also my understanding >> with >> > regards to documentation. >> > [Tue Jun 17 21:11:51 2014] <lmcgibbn>: Can we please agree that the >> > canonical source code for Usergrid (incubating) is at ASF git-wip >> > [Tue Jun 17 21:11:52 2014] <lmcgibbn>: ? >> > [Tue Jun 17 21:12:13 2014] <snoopdave>: yes, we agree on that >> > [Tue Jun 17 21:12:18 2014] <lmcgibbn>: OK >> > [Tue Jun 17 21:12:21 2014] <lmcgibbn>: One hurdle down >> > [Tue Jun 17 21:12:32 2014] <Humbedooh>: so I can send you a patch >> outside >> > of github then? >> > [Tue Jun 17 21:12:38 2014] <snoopdave>: policy says all releases must be >> > cut from canonical repo, and htat is what we do >> > [Tue Jun 17 21:13:09 2014] <lmcgibbn>: All patches are either on the ASF >> > Jira (which means they are ASLv2.0 licensed unless stated otherwise) or >> > [Tue Jun 17 21:13:16 2014] <lmcgibbn>: as per the workflow >> > [Tue Jun 17 21:13:31 2014] <lmcgibbn>: submitted as pull requests to the >> > mirrior at usergrid/usergrid >> > [Tue Jun 17 21:13:43 2014] <snoopdave>: humbedooh: we can accept patches >> > via JIRA but we urge people to use GitHub as that is our code review >> system >> > [Tue Jun 17 21:14:05 2014] <lmcgibbn>: Once (and if) code is merge there >> > it is merged back into ASF git-wip which is the canonical source >> > [Tue Jun 17 21:14:35 2014] <lmcgibbn>: The intermediate >> usergrid/usergrid >> > is merely for the convenience of the Usergrid project committers >> > [Tue Jun 17 21:14:56 2014] <lmcgibbn>: and for the convenience of the >> > contributors who may wish to become involved with developing Usergrid >> > [Tue Jun 17 21:15:06 2014] <toddnine>: Correct >> > [Tue Jun 17 21:15:26 2014] <lmcgibbn>: From a mentor point of view this >> > seems entirely logical to me. >> > [Tue Jun 17 21:15:26 2014] <rockerst_>: yes >> > [Tue Jun 17 21:15:29 2014] <Humbedooh>: if (IF) the canonical source for >> > the project is on git-wip and you put up a big yellow sign saying "This >> > GitHub repo is not the canonical source, please see....", then I don't >> see >> > a problem per se >> > [Tue Jun 17 21:15:43 2014] <lmcgibbn>: From a contributor point of view >> > this has been very convenient for me. >> > [Tue Jun 17 21:16:00 2014] <lmcgibbn>: OK >> > [Tue Jun 17 21:16:02 2014] <Humbedooh>: but if people are led to believe >> > that the github repo IS the canonical source, and you do nothing about >> it, >> > then it's your behinds on the line personally, and not the ASF >> > [Tue Jun 17 21:16:06 2014] <lmcgibbn>: We've made serious progress here. >> > [Tue Jun 17 21:16:13 2014] <lmcgibbn>: I mean serious progress. >> > [Tue Jun 17 21:16:17 2014] <rockerst_>: We would be happy to put up a >> > yellow sign >> > [Tue Jun 17 21:16:36 2014] <lmcgibbn>: As a mentor of this project, I >> feel >> > that this is a major step forward for the community (dev's especially). >> > [Tue Jun 17 21:16:53 2014] <lmcgibbn>: A lot of hard work has gone in to >> > Usergrid and I am really looking forward to sorting this out. >> > [Tue Jun 17 21:16:59 2014] <lmcgibbn>: Can we please take some actions? >> > [Tue Jun 17 21:17:50 2014] <lmcgibbn>: 1. All previous bullet points >> > regarding setup of mechanism that causes every pull request to be >> recorded >> > on the project commits list as emails. >> > [Tue Jun 17 21:17:56 2014] <rockerst_>: i think we still don't know >> > specifically who is objecting to the workflow >> > [Tue Jun 17 21:18:14 2014] <rockerst_>: lewis brought it up, but didn't >> > respond when we asked who is objecting >> > [Tue Jun 17 21:18:17 2014] <snoopdave>: and what specific objections >> they >> > have >> > [Tue Jun 17 21:18:31 2014] <lmcgibbn>: rockerst_: As far as this meeting >> > goes I think that no-one is objecting. >> > [Tue Jun 17 21:19:13 2014] <lmcgibbn>: I feel we have made progress on >> > clarifying what we (Usergrid) want to achieve... >> > [Tue Jun 17 21:19:21 2014] <lmcgibbn>: It seems like we've been able to >> do >> > that. >> > [Tue Jun 17 21:19:37 2014] <rockerst_>: lmcgibbn: agree >> > [Tue Jun 17 21:19:44 2014] <lmcgibbn>: OK >> > [Tue Jun 17 21:19:48 2014] <lmcgibbn>: So I propose the following >> > [Tue Jun 17 21:19:52 2014] <lmcgibbn>: In addition to 1. above >> > [Tue Jun 17 21:20:01 2014] <lmcgibbn>: we also address Humbedooh >> suggestion >> > [Tue Jun 17 21:20:13 2014] <rockerst_>: yellow sign? >> > [Tue Jun 17 21:20:17 2014] <lmcgibbn>: +1 >> > [Tue Jun 17 21:20:23 2014] <Humbedooh>: well...it could be blue, if you >> > like blue ;-) >> > [Tue Jun 17 21:20:30 2014] <lmcgibbn>: :) >> > [Tue Jun 17 21:20:48 2014] <toddnine>: +1 >> > [Tue Jun 17 21:21:31 2014] <lmcgibbn>: I would also like to finally >> > clarify that script on snoopdave 's account has been deactivated based >> on >> > recent actions taken by Infra leading up to this IRC meeting. >> > [Tue Jun 17 21:21:42 2014] <lmcgibbn>: snoopdave: can you please clarify >> > for the record? >> > [Tue Jun 17 21:22:00 2014] <snoopdave>: yes, my script was disabled >> > [Tue Jun 17 21:22:04 2014] <lmcgibbn>: OK >> > [Tue Jun 17 21:22:24 2014] <snoopdave>: when I need to push from GH to >> ASF >> > Git I do it manually >> > [Tue Jun 17 21:22:42 2014] <lmcgibbn>: So by the looks and sounds of it >> > the workflow as stated on the Usergrid (incubating) wiki can continue >> once >> > we have addressed the above actions. >> > [Tue Jun 17 21:23:10 2014] <rockerst_>: i can update the readme file in >> > the root of the project to convey the information about the canonical >> source >> > [Tue Jun 17 21:23:22 2014] <lmcgibbn>: +1 rockerst_ thank you >> > [Tue Jun 17 21:23:30 2014] <lmcgibbn>: Does anyone have anything to add? >> > [Tue Jun 17 21:24:32 2014] <snoopdave>: I think that is all we can do: >> > address specific problems with our process: 1) indicate our repo is not >> the >> > canonical repo, 2) setup email notifications for PRs and 3) figure how >> to >> > automate the sync >> > [Tue Jun 17 21:24:44 2014] <rockerst_>: exactly >> > [Tue Jun 17 21:24:54 2014] <lmcgibbn>: Sounds great. >> > [Tue Jun 17 21:24:55 2014] <rockerst_>: i think we are done here >> > [Tue Jun 17 21:25:01 2014] <lmcgibbn>: RE: 3 >> > [Tue Jun 17 21:25:07 2014] <lmcgibbn>: any immediate ideas? >> > [Tue Jun 17 21:25:09 2014] <snoopdave>: thanks Humbedooh >> > [Tue Jun 17 21:25:40 2014] <rockerst_>: lmcgibbn: jfarrell: Humbedooh: >> > thank you for helping out :) >> > [Tue Jun 17 21:25:52 2014] <lmcgibbn>: Humbedooh: jfarrell thank you for >> > your time >> > [Tue Jun 17 21:25:59 2014] <Humbedooh>: re 3, I'd hold off till the >> policy >> > review is done >> > [Tue Jun 17 21:26:15 2014] <lmcgibbn>: Humbedooh: +1 >> > [Tue Jun 17 21:26:30 2014] <lmcgibbn>: Humbedooh: can you please adjourn >> > this meeting when you are ready. >> > [Tue Jun 17 21:26:34 2014] <lmcgibbn>: Thank you for your time. >> > [Tue Jun 17 21:26:54 2014] <lmcgibbn>: BTW people, #usergrid is much >> > better when people are actually here. >> > [Tue Jun 17 21:26:58 2014] <Humbedooh>: ASFBot: karma add lmcgibbn 3 >> > [Tue Jun 17 21:27:03 2014] <Humbedooh>: lmcgibbn: you can do it >> yourself ;) >> > [Tue Jun 17 21:27:08 2014] <lmcgibbn>: recently it has been me and >> jfarrell >> > [Tue Jun 17 21:27:11 2014] <lmcgibbn>: Thank you >> > [Tue Jun 17 21:28:26 2014] <Humbedooh>: ASFBot: meeting end >> > >> > >> > Meeting ended at Tue Jun 17 21:28:26 2014 >> > >> >
